Su1532

Outcomes After Transhiatal and Transthoracic Esophagectomy for EsophagealCancerChristopher S. Davis, Eileen Bock, Kirstyn E. Brownson, Cynthia Weber, P. MarcoFisichella, Margo Shoup, Gerard V. Aranha

Background: Controversy persists as to the preferred operative approach to esophagealcancer. Therefore, we investigated the peri-operative, short-term, and mid-term outcomesbetween transhiatal esophagectomy (THE) and transthoracic esophagectomy (TTE) at ourinstitution. Methods: We conducted a retrospective review of 114 patients who had under-gone esophagectomy for esophageal cancer, in our tertiary care center. Among those patientswho underwent THE or TTE we compared: a) clinical characteristics; b) pathologic findings;and c) outcomes. Parametric and non-parametric tests of significance were performed, andsurvival was determined by Kaplan-Meier analysis. Results: We identified 32 patients whounderwent THE, and 82 patients who underwent TTE. Age, gender, race/ethnicity, alcoholand tobacco use, weight loss and body mass index at the time of surgery, operative risk,chemoradiation regimen, tumor stage, and pathologic findings were similar between groups.Those who underwent THE had a greater intra-operative blood loss (p=0.003), requiredmore intra-operative blood transfusions (p<0.0001), spent a longer time on the ventilator(p<0.0001) and in the intensive care unit (p=0.002), and had a higher 30-day mortality(p=0.023). Likewise, those who underwent THE had a greater prevalence of post-operativevocal cord dysfunction (17% vs 3%, p=0.027) and anastomotic leak (29% vs. 1%, p<0.0001).Compared to THE, patients undergoing TTE had a greater number of lymph nodes sampled(mean 13.0 vs 13.6, respectively) and frequency of lymph nodes positive for carcinoma(29% vs 35%, respectively). Finally, survival at 3-years was significantly less after THE thanafter TTE (26% vs. 53%, p=0.035), as was overall 5-year survival (p=0.039) (Figure 1).Conclusions: These data demonstrate a short-term survival advantage and lower morbidityof TTE as compared to THE at our institution. We speculate that the higher morbidity afterTHE may account for the worse outcomes associated with this approach.

Su1533

Venous Thromboembolism in Patients Receiving Neoadjuvant Chemotherapyfor Esophagogastric CarcinomaDavid Bowrey, Achal Khanna, Alex M. Reece-Smith, Anne Thomas, Simon Parsons

Background: The association between venous thromboembolism and chemotherapy foresophagogastric cancer is well known in patients treated with palliative intent. Whether thisrisk extends to the neoadjuvant and perioperative setting is unclear. Methods: Retrospectiveinterrogation of databases of patients receiving perioperative chemotherapy for potentiallycurative intent at the Leicester (2006-2011) and Nottingham (2004-2011) esophagogastriccancer centres. Results: Thromboembolic events were diagnosed in 42 of 384 patients (11%),16 (4%) at presentation, 14 (4%) during neodjuvant chemotherapy and 12 (3%) in thepostoperative period. By site these comprised catheter-related axillary vein thrombosis in 6patients, deep venous thrombosis in 16 patients and pulmonary embolism in 16 patients.All of the pulmonary emboli were incidental findings on staging CT imaging. There was nocorrelation between the risk of thromboembolism and chemotherapy regimen. Seven of the42 patients (17%) who developed thromboembolism did not proceed to surgery because

S-1057 SSAT Abstracts

of deterioration in performance status. Thromboembolic disease resulted in a non-significantincrease in the interval between chemotherapy and surgery, but did not influence eitherlength of hospital stay or survival. Conclusions: Eleven percent of patients treated withpotentially curative intent will develop venous thromboembolism. This adverse event canoccur at any time during the patient journey. In contrast to the commonly held view, thisdid not translate into a poorer prognosis.

Su1535

Methylene Blue (MB) Test Versus Contrast Study (CS) in the Detection ofAnastomotic Leak Following Oesogephactomy: A Prospective Study of 58PatientsAdriana Rotundo, Geoffrey Roberts, Francesco Pata, Geoff Pratt, Michael Harvey, CheukBong Tang, Sritharan S. Kadirkamanathan

Background: Anastomotic leak is a serious complication following oesophagectomy. It isassociated with considerable morbidity and mortality. The aim of our study was to comparethe accuracy of MB and CS (Gastrografin) in detecting anastomotic leaks after Ivor-Lewisoesophagectomy. Methods: The study included 58 patients who underwent laparoscopicassisted Ivor-Lewis oesophagectomy from September 2009 to November 2011. All patientshad intra-thoracic oesophago-gastric anastomosis, end to side using an endoscopic circularstapler (CDH © Ethicon Endo-Surgery, Inc.2010). The integrity of the anastomosis waschecked on day 5 using both MB and CS. 100ml of Gastrografin was used in the CS whichwas performed by an experienced oesophageal radiologist. 10ml of MB diluted in 200ml ofwater was given orally to test the anastomosis. The CS was performed before the MB testand was reviewed by the radiologist who was blinded to the results of MB. MB test wasconsidered positive when the dye was seen in the chest drain in less than 30 minutes. Theleak was considered clinically significant if there was evidence of sepsis. Chi square test wasused to assess the difference between the two investigations Results: There were 37 malesand 21 females with median age of 65 (range 43-78). Anastomotic leaks was diagnosed in6 patients (10.3%). In 4 cases the leak was considered clinically significant (7%). MB detectedall 4 significant leaks. CS detected 5 leaks, 3 clinical and 2 non-clinical, but was reportedas a normal study in 1 of the clinically significant leak. All patients recovered with conservativemanagement. There was no significant difference between MB and CS in diagnosing anastom-otic leaks (p=ns) Conclusion: Our study shows no difference between CS and MB in detectinganastomtic leaks. MB might be a more convenient investigation and could be used in a wardsetting without the need for radiology. It could well form part of the strategy of enhancedrecovery after surgery (ERAS) programme following oesophageal surgery.

Su1536

Defining the Learning Curve for Robotic-Assisted EsophagogastrectomyJonathan M. Hernandez, Jill Weber, Khaldoun Almhanna, Sarah Hoffe, Ravi Shridhar,Richard Karl, Ken L. Meredith

Introduction: The expansion of robotic-assisted surgery is occurring quickly, though littleis generally known about the “learning curve” for the technology with utilization for complexesophageal procedures. The purpose of this study is to define the learning curve for robotic-assisted esophagogastrectomy with respect to operative time, conversion rates, and patientsafety. Methods: We have prospectively followed all patients undergoing robotic-assistedesophagogastrectomy and compared operations performed at our institutions by a singlesurgeon in successive cohorts of 10 patients. Our measures of proficiency included: operativetimes, conversion rates, and complications. Results: Fifty-two patients (41 (78.8%) male:11 (22.2%) female) of mean age 66.2 ± 8.8 years underwent robotic-assisted esophagogastrec-tomies for malignant esophageal disease. Neoadjuvant chemoradiation was administered to 35(67.3%) patients. A significant reduction in operative times (p<0.005) following completion of20 procedures was identified (514 ± 106 vs. 397 ± 71.9). No significant reduction in thenumber of procedures requiring conversions to open operations was observed. Complicationrates were low, and not significantly different between any 10-patient cohort, although nocomplications occurred in the final 10-patient cohort (Figure 1). However the frequency ofcomplications decreased significantly after 28 cases: 9 (32.1%) vs 3 (12.5%) p=0.04. Therewere no in hospital mortalities. Conclusions: For surgeons proficient in performing minimallyinvasive esophagogastrectomies, the learning curve for a robotic-assisted procedure appearsto begin near proficiency after 20 cases. However this may be increased in surgeons trans-itioning from an open approach. Operative complications and conversions were infrequentand unchanged across successive 10-patient cohorts and appear to be less then smallerpreviously published series. In addition, there is a decrease in frequency of complicationsafter 28 cases.

Su1537

Risk Factors for Postoperative Mortality After General Surgery in 231 PatientsWith Liver CirrhosisFrank Makowiec, Hans-Christian Spangenberg, Tobias Keck, Ulrich T. Hopt, Hannes P.Neeff

Postoperative mortality rates after surgery in patients with liver cirrhosis are high. Riskfactors for mortality may help planning therapy in those high risk patients. We, therefore,evaluated/updated potential risk factors (including Child- andMELD-scores) for perioperativemortality after more than 200 operations in patients with cirrhosis performed during thelast decade. Methods: Since 2001 231 various general surgical procedures (80% intraabdomi-nal, 20% abdominal wall) were performed in patients with liver cirrhosis (38% emergent).Cirrhosis was classified according to Child (41% A; 38% B, 21% C) and MELD-score (median11). Procedures were subclassified as major (laparotomy with resection) or minor (abdominalwall, ‘minor' laparotomy, laparoscopy). Univariate and multivariate (binary logistic regres-sion) analysis was undertaken to identify risk factors for mortality. Multivariate analysis wasperformed in different models to exclude collinearity due to overlapping parameters (Child,MELD, laboratory values). Results: Overall postoperative mortality was 17%. In univariate

SS

AT

Ab

stra

cts